Artem

Add support for TF saved model / TF hub models

1

In TF it's common practise to use `tf.saved_model` format, for (de)serialising trained models. This format has however a bit other APIs than `tf.keras.Model`. It would be great to have Quantus...

fixtures used lazily are not detected

3

Hi, after I run ` pytest --dead-fixtures` I get an output e.g.  To double check if ii is so, I run `grep -r 'almost_uniform_1d' tests` and get . So,...

From top to down the views generated by: - PyCharm - Colab - JupyterLab - GitHub preview     Ideally, we want them all to be rendered in...

### Description Added a notebook showing proposal of how quantitative evaluation of NLP models could look in future ### Implemented changes added Getting_Started_with_NLP.ipynb ### Minimum acceptance criteria ???

GlueModel from example returns map object instead of list

2

Hi, as you can see on the 1st screenshot, `GlueModel` from `lit_nlp.examples.models.glue_models` returns map object, instead of intended list. After executing `map_result = nn.predict_minibatch([{'sentence1': 'Hello there'}])` `list_result = list(map_result)`, the...

Hi team `tensorflow-model-optimization`, I was applying clustering and pruning to a model based on Bert encoder from `tensorlfow-models-official`, and noticed that there is no API for registering custom layers, so...

Dropout + `nn.jit`

3

Hi there, I was following this guide, [flax.linen.Dropout](https://flax.readthedocs.io/en/latest/guides/dropout.html#more-flax-examples-that-use-module-make-rng). Then I decided to add `nn.jit`, and started getting `MyModel.__call__() missing 1 required positional argument: 'training'`, even though I passed it. ###...

Multi-label classification support

1

### Description of the problem - afaik so far, Quantus assumes categorical encoding of `y_batch`, where each input data-point is assigned to exactly 1 label. ### Description of a solution...

### Description of the problem - A lot of metrics do computations on a single instance. - They often could be computed in batches. ### Description of a solution Apply...